| 00:25:41 | * | q66 quit (Remote host closed the connection) |
| 01:26:30 | * | Anaphaxeton joined #nimrod |
| 03:07:03 | * | XAMPP quit (Read error: Connection reset by peer) |
| 05:07:12 | * | fowl joined #nimrod |
| 05:07:43 | * | fowl_ joined #nimrod |
| 05:08:28 | * | fowl quit (Client Quit) |
| 05:08:28 | * | fowl_ quit (Client Quit) |
| 05:08:36 | * | fowl joined #nimrod |
| 05:08:43 | * | fowl quit (Client Quit) |
| 05:09:55 | * | fowl joined #nimrod |
| 05:35:09 | * | XAMPP joined #nimrod |
| 05:35:09 | * | XAMPP quit (Changing host) |
| 05:35:09 | * | XAMPP joined #nimrod |
| 05:46:42 | fowl | Araq: no idea where to get libffi binaries for windows |
| 05:47:12 | fowl | the first page of google didnt solve my problem and i dont feel like compiling make and shit because apparently i dont have that either |
| 06:33:01 | * | FreeArtMan joined #nimrod |
| 07:14:04 | * | XAMPP quit (Quit: My code has no bug's, just random features) |
| 07:21:17 | * | q66 joined #nimrod |
| 07:33:04 | * | gour joined #nimrod |
| 07:43:14 | * | fowl quit (Remote host closed the connection) |
| 07:53:59 | * | fowl joined #nimrod |
| 08:30:52 | * | Araq_ joined #nimrod |
| 08:34:21 | * | Araq_ quit (Client Quit) |
| 08:43:25 | * | q66 quit (Remote host closed the connection) |
| 09:08:03 | * | gour quit (Disconnected by services) |
| 09:08:04 | * | gour_ joined #nimrod |
| 09:19:15 | * | gour_ is now known as gour |
| 09:20:44 | * | Anaphaxeton quit (Ping timeout: 252 seconds) |
| 10:36:43 | * | XAMPP joined #nimrod |
| 12:14:22 | Zor | fowl: you can build libffi reasonably easily on cygwin, mingw32, mingw-w64 |
| 12:14:47 | Zor | should build out of the box on all 3 |
| 12:39:37 | * | Anaphaxeton joined #nimrod |
| 12:52:41 | * | Anaphaxeton quit (Quit: Αποχώρησε) |
| 13:25:10 | * | fowl quit (Quit: Leaving) |
| 13:31:01 | * | q66 joined #nimrod |
| 13:43:13 | * | XAMPP-8 joined #nimrod |
| 13:44:50 | * | XAMPP-8 quit (Client Quit) |
| 15:21:37 | * | FreeArtMan quit (Ping timeout: 248 seconds) |
| 15:43:26 | gour | Araq: thank you for your answer in the forum, but replacing #define with #def still does not work on my minimal example - see http://pastebin.com/5qUM2pbQ |
| 15:45:06 | gour | iow. it fails on function declaration ext_def(int32) swe_heliacal_ut |
| 15:45:25 | gour | adding/remolving ';' does not help |
| 17:25:23 | reactormonk | Araq, btw, close or accept those pull requests imo |
| 17:32:56 | * | Anaphaxeton joined #nimrod |
| 17:38:07 | * | FreeArtMan joined #nimrod |
| 17:47:53 | * | Anaphaxeton quit (Remote host closed the connection) |
| 17:48:30 | * | Anaphaxeton joined #nimrod |
| 17:56:02 | Araq | ping gour |
| 18:00:35 | * | filwit joined #nimrod |
| 18:00:46 | filwit | hi guys |
| 18:02:04 | filwit | Araq, dom96, you around? |
| 18:13:39 | dom96 | filwit: I am now ;) |
| 18:14:19 | filwit | hey dom96 :) |
| 18:14:27 | dom96 | hey filwit :D |
| 18:14:30 | dom96 | What's up? |
| 18:14:40 | filwit | so, I know the guy who runs: http://worldofgnome.org/ |
| 18:14:48 | filwit | and he's going to make me an editor |
| 18:14:58 | filwit | and I was thinking we could do a Nimrod article |
| 18:15:03 | filwit | when we launch the new site |
| 18:15:04 | dom96 | cool! |
| 18:15:10 | dom96 | great Idea. |
| 18:15:16 | filwit | will have to be Gnome centered |
| 18:15:24 | filwit | so GTK bindings, etc |
| 18:15:28 | dom96 | That might be a challenge |
| 18:15:34 | dom96 | You could mention Aporia though |
| 18:15:48 | dom96 | It's already a pretty good alternative to gedit |
| 18:16:00 | dom96 | Will need to release before that though |
| 18:16:18 | filwit | well I'm thinking a tutorial on building a GTK app in Nimrod |
| 18:17:15 | filwit | one sec |
| 18:17:28 | dom96 | The GTK wrappers are not very Nimrodic yet though :\ |
| 18:17:58 | filwit | dom96: we'll only use GTK as a way to sneak in cool Nimrod language features show offs |
| 18:18:16 | filwit | Python is used in Gnome a lot |
| 18:18:37 | filwit | so Nimrod is kinda like the "Fast/Safe version of Python" |
| 18:18:46 | filwit | (at least that's going to be people first impressions) |
| 18:21:03 | dom96 | filwit: how big is world of gnome? |
| 18:21:13 | filwit | ... sorry, talking to someone ATM |
| 18:21:29 | filwit | um.. take a look at the traffic: http://wogue.org/piwik/index.php?module=CoreHome&action=index&idSite=1&period=day&date=yesterday |
| 18:21:32 | filwit | it's not huge |
| 18:21:36 | filwit | by any means |
| 18:21:49 | filwit | but 1,400-2,000 hits a day isn't bad |
| 18:21:59 | dom96 | indeed. |
| 18:22:55 | filwit | it will bring a lot of eyes to Nimrod, and if we corrolate it with the new site, we can get quite a few good "first impressions" |
| 18:24:45 | * | Anaphaxeton quit (Quit: Αποχώρησε) |
| 18:25:04 | Araq | filwit: excellent, thanks a lot |
| 18:25:25 | filwit | np, my pleasure |
| 18:25:46 | filwit | like i said, i would like to see Nimrod get bigger as well |
| 18:26:41 | filwit | that way.. my complaints will be echoed in the larger voice of the populous, and you will have to recognize the correctness of my arguments :P |
| 18:27:23 | filwit | (in actuality, I think I'm probably the only person that cares so much) |
| 18:29:03 | filwit | anyways, since you way it will be around 1-2 months for the next version of Nimrod to be released |
| 18:29:08 | dom96 | hrm, is there a big blog which focuses strictly on programming language news? |
| 18:29:14 | filwit | we should start writing an article now |
| 18:29:22 | Araq | dom96: lambda the ultimate? |
| 18:29:25 | filwit | and get it ready so we can do a "marketing push" on that date |
| 18:29:56 | dom96 | Araq: hrm, true. |
| 18:29:57 | filwit | dom96: probably, but not that I can get an article into :P |
| 18:30:40 | dom96 | We should submit a link to lambda the ultimate simultaneously :P |
| 18:30:58 | filwit | yes good idea |
| 18:30:59 | Araq | I guess I could get an article to LTU |
| 18:31:24 | filwit | and if you guys use Twitter (i don't) then announcements there, and Reddit, etc |
| 18:31:48 | dom96 | yeah. |
| 18:32:04 | reactormonk | dom96, did you implement nimrod serve yet? |
| 18:32:13 | dom96 | reactormonk: in aporia, no. |
| 18:32:26 | reactormonk | dom96, do you have any docs? |
| 18:32:32 | dom96 | reactormonk: nope, sorry. |
| 18:32:39 | Araq | reactormonk: nobody has :-) |
| 18:32:45 | reactormonk | Araq, see issue... |
| 18:32:51 | reactormonk | Araq, can you assign that to zah? |
| 18:33:39 | dom96 | reactormonk: You could take a look at the source code if you're desperate ;) |
| 18:33:51 | Araq | reactormonk: done |
| 18:34:09 | filwit | also, I was going to say that the web-site should have better pages than it does now. For instance "Community" should actually talk the community, goals, and maybe an exert about Araq and other contributors, etc |
| 18:34:30 | reactormonk | dom96, sure |
| 18:34:44 | filwit | the maybe have a few more code examples in the docs page (like the one on the home page) |
| 18:34:47 | dom96 | filwit: I agree. |
| 18:34:47 | reactormonk | filwit, and use bootstrap? |
| 18:35:11 | filwit | reactormonk: what do you mean? |
| 18:35:17 | reactormonk | filwit, twitter bootstrap |
| 18:35:26 | * | Anaphaxeton joined #nimrod |
| 18:35:36 | dom96 | reactormonk: use twitter bootstrap for what? |
| 18:35:39 | filwit | reactormonk: i know what bootstrap is, i just don't know what context you're talking about |
| 18:35:57 | filwit | reactormonk: do you mean a tutorial on how to build Nimrod? |
| 18:36:27 | reactormonk | dom96, for the webpage |
| 18:36:42 | dom96 | reactormonk: what benefits would that bring? |
| 18:36:49 | reactormonk | dom96, looks fancy |
| 18:36:59 | dom96 | reactormonk: I guess you haven't seen filwit's design? |
| 18:37:04 | reactormonk | dom96, nope |
| 18:37:22 | filwit | reactormonk: http://reign-studios.com/nimrod/web/ |
| 18:37:29 | dom96 | Also, I dislike twitter bootstrap. |
| 18:37:41 | dom96 | There are too many websites which use it |
| 18:37:45 | reactormonk | filwit, woah. Fancy. |
| 18:37:47 | dom96 | and I find it repulsive now. |
| 18:37:49 | filwit | :) |
| 18:37:56 | reactormonk | filwit, but somehow the character spacing is a bit too wide. |
| 18:38:19 | reactormonk | filwit, and it was 'Rumpf' iirc |
| 18:38:29 | filwit | i am adding a bit of extra spacing, but i like it. What browser/OS are you on? |
| 18:38:39 | reactormonk | chromium / Linux |
| 18:38:51 | filwit | reactormonk: plus, we can fiddle with the CSS and decide what works best |
| 18:39:10 | reactormonk | filwit, http://ompldr.org/vaGl1NQ |
| 18:40:02 | filwit | reactormonk: okay, I'm using the same setup and I don't mind the spacing :) |
| 18:40:15 | filwit | reactormonk: ps. No AdBlock? |
| 18:40:17 | reactormonk | filwit, ... are the tabs supposed to work? |
| 18:40:29 | Araq | reactormonk: it's a *design* |
| 18:40:31 | reactormonk | filwit, sure, see that little icon next to the star? |
| 18:40:36 | filwit | reactormonk: not yet, it's just the template |
| 18:41:01 | Araq | it's also "welcome" and not "welcom", but who cares |
| 18:41:03 | filwit | reactormonk: oh, you're shows up differently than mine |
| 18:41:04 | reactormonk | filwit, ok. |
| 18:41:06 | Araq | it's a design |
| 18:41:27 | filwit | oh, whoops, misspelled that |
| 18:42:52 | filwit | fixed |
| 18:43:04 | Araq | filwit: the problem with better docs is that nobody reads them anyway |
| 18:43:05 | filwit | idk how I turned "Welcome" into "Wecom" |
| 18:43:06 | filwit | lol |
| 18:43:19 | filwit | Araq: i didn't mean changed the docs |
| 18:43:29 | Araq | seriously, just look at the latest thread on the forum |
| 18:43:36 | Araq | it's like talking to a wall |
| 18:43:39 | filwit | only give a few "quick examples" on the docs page, under the links to the official docs site |
| 18:44:06 | filwit | like the little code example on the homepage |
| 18:44:19 | filwit | Araq: I've never browsed the forums before |
| 18:44:29 | filwit | Araq: cause they're so ugly |
| 18:44:36 | Araq | :-( |
| 18:44:43 | filwit | lol |
| 18:44:52 | Araq | I like the design |
| 18:44:55 | reactormonk | filwit, hm. letter-spacing: 0.9px and 1px gives widely different results |
| 18:45:19 | filwit | reactormonk: it's rounded down. 0.9px is 0px |
| 18:45:25 | reactormonk | oh. |
| 18:45:32 | reactormonk | Anyway, I prefer http://i.imgur.com/SnaWCSj.png ;-) |
| 18:45:33 | filwit | you can't have 0.9px lol |
| 18:45:36 | filwit | they're pixels |
| 18:45:54 | reactormonk | filwit, hacking without brainz - works sometimes. |
| 18:46:22 | filwit | :) |
| 18:47:06 | dom96 | filwit: :( |
| 18:47:07 | reactormonk | letter spacing in the quote is fine imo. |
| 18:47:08 | filwit | let's just take a community vote later about the letter-spacing, or do whatever Araq likes best |
| 18:47:12 | dom96 | That's my crappy design </3 |
| 18:47:29 | filwit | dom96: :O sorry |
| 18:47:41 | dom96 | lol it's ok |
| 18:47:56 | filwit | it's just isn't exactly what a modern forum should look like, IMO |
| 18:48:12 | dom96 | yeah, I agree. |
| 18:48:18 | dom96 | I didn't have time to design it well |
| 18:48:25 | filwit | yes i understand |
| 18:48:50 | dom96 | but you can design it too if you want ;) |
| 18:49:01 | filwit | getting things done is more important than nit-picking all the aesthetic details... at least to everyone but me |
| 18:49:38 | Araq | filwit: the real test is whether the design is still bearable within a year |
| 18:49:42 | filwit | dom96: we can redesign the forum later, maybe just house your design (with some CSS mods) in the new sight layout or something |
| 18:50:20 | filwit | Araq: I have a nagging feeling that one day you're going to wake up and hate the new website, and then take it down, lol |
| 18:50:31 | filwit | Araq: which is fine, it is your website afterall |
| 18:50:40 | Araq | that's a real danger ;-) |
| 18:50:58 | Araq | I'm happy I still like the current website and forum |
| 18:51:15 | Araq | it's not super hot, but it ages slowly :P |
| 18:51:29 | Araq | ymmv of course |
| 18:51:52 | filwit | well even if you like it a lot, I think having a "fancy looking" website with help Nimrod attract attention |
| 18:52:31 | filwit | ymmv? |
| 18:53:56 | filwit | people often underestimate the power of good advertising. I've watched companies double and triple their customers overnight due to a new ad graphics and slogan |
| 18:54:02 | filwit | without even advertising more |
| 18:54:18 | dom96 | I completely agree with you filwit |
| 18:54:55 | dom96 | It's a subconscious thing I think. |
| 18:54:58 | filwit | they just put a new catch-phrase on their existing ad-spots and change the graphics to look better and stand out more (compared to other competitors), and bam, 2-3 times more interest |
| 18:55:07 | filwit | dom96: yeah exactly |
| 18:55:20 | filwit | dom96: most advertising isn't even about saying anything |
| 18:56:36 | filwit | dom96: it's just about association. You're more likely to buy a product off the shelves, if you've seen it on TV, and the ad left you with a positive feeling. That emotional memory with the images sticks, even if you're not conscious of it. |
| 18:57:20 | filwit | that's why all those stupid skiddles and snickers bars ads work |
| 18:59:26 | dom96 | indeed |
| 18:59:41 | filwit | anyways folks, I need to get back to things. Just wanted to let you know about the article possibility |
| 19:00:01 | filwit | I will want help writing that, and coming up with good material |
| 19:00:30 | Araq | excellent |
| 19:01:33 | filwit | oh, ps. Kudos on that "term rewriting macros" feature |
| 19:01:50 | filwit | i saw that cause of Doms post on the Math-D article |
| 19:01:54 | filwit | that's really handy |
| 19:02:00 | Araq | :-) |
| 19:02:10 | filwit | (though, if misused, could introduce really hard-to-find bugs) |
| 19:02:35 | Araq | yeah but it's never really been tried before |
| 19:02:36 | dom96 | yeah :D |
| 19:02:40 | dom96 | I knew people would see it :D |
| 19:02:45 | filwit | it could be really great for example, translating game-script code (written by "noobs") into simd-optimized code |
| 19:02:51 | Araq | yep. |
| 19:03:09 | filwit | ps. what's the situation with SIMD btw? |
| 19:03:17 | filwit | and work done on that front? |
| 19:03:26 | filwit | i've learned a lot about that recently |
| 19:03:35 | filwit | not an expert or anything |
| 19:03:41 | Araq | we had a guy that wanted to do it |
| 19:03:44 | filwit | but maybe i can hack it into the compile with some help |
| 19:03:47 | Araq | but apparently he is too busy |
| 19:04:19 | Araq | filwit: you would have my full support for that |
| 19:04:31 | Araq | it's easily wrapped with the existing FFI |
| 19:04:50 | Araq | and as you said, TR macros for sugar |
| 19:04:52 | filwit | Araq: yes dom96 mentioned that the other day |
| 19:05:06 | filwit | Araq: i will look into the FFI lib support you added |
| 19:06:07 | filwit | Araq: but i don't really think TR macros should be used soley. Maybe there should be a compile flag with introduces extra TRMacro modules into the compile |
| 19:06:19 | filwit | but i think there should be a SIMD-optimized Vector type |
| 19:06:56 | filwit | since often, the only way to get good performance is to limit how you access the vector components, and encourage using vectors "as a single number" type of thing |
| 19:07:38 | Araq | filwit: I've seen benchmarks where Nimrod outperforms C where the C uses SIMD intristics |
| 19:07:42 | * | zahary joined #nimrod |
| 19:07:44 | Araq | and Nimrod simple arrays |
| 19:07:44 | filwit | i will probably just copy Manu's std.simd (D) lib (which uses C's ia_xxx functions) |
| 19:08:09 | dom96 | hrm, what is SIMD? |
| 19:08:12 | filwit | Araq: well.. Nimrod compiles to C, and GCC auto-vectorizes |
| 19:08:20 | Araq | GCC's auto vectorization is not bad these days |
| 19:08:29 | Araq | yes, that's my point |
| 19:09:00 | filwit | Araq: yes, but it's easily beaten with specialized vector types when used right. |
| 19:09:06 | filwit | Araq: i've done tests |
| 19:09:23 | Araq | filwit: the C was code from the language shootout iirc |
| 19:09:36 | Araq | but sure, go ahead I don't mind |
| 19:10:02 | Araq | btw TR macros basically require you to have version that doesn't require them |
| 19:10:10 | filwit | Araq: the problem is, things like Matrix multiplication is to complicated for a auto-vectorizor to really figure out the best way to do things (at least today). There are a lot of tricks, and even more with different SIMD instructions |
| 19:10:37 | Araq | matrix multiplication is a simple pattern, easily hard coded in a compiler |
| 19:10:41 | filwit | Araq: yes, I think TR macros are definitely worth looking into. |
| 19:11:25 | filwit | Araq: there's actually a few different ways for the code to happen, and so you have to analyze for more than just one pattern |
| 19:11:42 | filwit | that the code can happen** |
| 19:11:46 | Araq | hi zahary, I got rid of the cellsets for the mark&sweep GC, but it doesn't work yet :-/ |
| 19:12:07 | Araq | filwit: sure, but you should test Intel's compiler |
| 19:12:25 | filwit | yes, I hear good things about ICC |
| 19:12:52 | filwit | but apparently it's only really good for Intel, and it's closed.. so whatever |
| 19:12:56 | filwit | i don't have much interest |
| 19:14:15 | filwit | dom96: SIMD = Single Instruction Multiple Data, it stands for special CPU hardware operations that work on a set of data usually used for Math & Game code (Vectors) |
| 19:14:42 | dom96 | ahh, from a quick skim read of the wiki article I think I get it. |
| 19:15:15 | filwit | dom96: basically, it just allows you to do 4 float operations in 1 operation |
| 19:15:21 | filwit | dom96: plus other tricks |
| 19:15:30 | Araq | filwit: even cooler would be to patch the C codegen so that it supports CUDA |
| 19:15:59 | filwit | don96: for example, you can "shuffle" the vector (rearange it's components) in 1 operation, which means you can compress some math operations even further |
| 19:16:38 | filwit | Araq: i would use OpenCL since it's a standard and cross platform (even though CUDA is better apparently) |
| 19:17:04 | Araq | yeah I guess you're right |
| 19:17:16 | * | FreeArtMan quit (Quit: rm -rf /) |
| 19:17:27 | filwit | Araq: plus, WebCL is coming out soon |
| 19:17:37 | Araq | omg |
| 19:17:57 | * | Araq despises anything web related |
| 19:17:59 | filwit | Araq: so in theory, you could use Nimrods JS backend + WebGL + WebCL bindings to write performant web apps |
| 19:18:11 | * | filwit agrees with Araq there |
| 19:18:37 | filwit | unfortunately, the web technology isn't going anywhere, at least not right now |
| 19:19:07 | filwit | i think eventually it will evolve into something more like Arch's repos, but it will take a long time for that to happen |
| 19:21:25 | filwit | dom96: one last note about SIMD: they're vital to Game Engines. SIMD can boost math performance (the major meat of Game code) by 4-10x in most cases. AAA game companies spend a lot of money in optimizing their core SIMD code. |
| 19:21:38 | filwit | dom96: that's why Nimrods TR macros look really cool |
| 19:22:15 | filwit | dom96: cause you can automate that process once, instead of paying someone tons of money to check every situation in the SIMD code |
| 19:23:46 | filwit | my basic idea about how to do SIMD in Nimrod will probably be a hybrid of a Vector type + TR macros, where the macros search for Vector patterns and fix common use-case mistakes, but the actual SIMD instructions are done through the Vector's procs |
| 19:24:20 | filwit | all the other "general math" cases are probably being optimized by GCC anyways. |
| 19:24:49 | filwit | anyways, i need to get back to work |
| 19:24:51 | filwit | later forlks |
| 19:24:56 | dom96 | oh wait |
| 19:25:04 | filwit | what's up? |
| 19:25:21 | dom96 | Perhaps docs could use a new design too ;) |
| 19:25:37 | Araq | don't you dare :P |
| 19:25:54 | filwit | i think the docs are fine the way they are |
| 19:25:54 | dom96 | Araq: You need to embrace change, change is good :P |
| 19:26:05 | filwit | they're easy to read, and very technical to begin with |
| 19:26:16 | Araq | ha! |
| 19:26:20 | dom96 | lol. |
| 19:26:26 | Araq | see, dom96, our docs rule |
| 19:26:28 | dom96 | Alright. |
| 19:26:41 | filwit | maybe one day, we can touch them up |
| 19:26:50 | filwit | the Forum is a bigger priority |
| 19:27:55 | filwit | and Aporia should have it's own page (not just github) |
| 19:28:11 | dom96 | yes, well actually. Aporia needs an icon first :) |
| 19:28:11 | filwit | basically, the three main "more links" buttons should take you to something nice looking |
| 19:28:16 | filwit | except Nimbuild |
| 19:28:20 | filwit | that's fine as-is |
| 19:28:29 | dom96 | hey, Nimbuild is already nice looking :P |
| 19:28:36 | filwit | yes i know |
| 19:28:49 | filwit | that's perfectly fine |
| 19:28:51 | dom96 | I need to add some docs to it though |
| 19:28:58 | dom96 | as people find it confusing |
| 19:29:52 | filwit | idk, most people who know what they're looking at will understand it |
| 19:30:06 | filwit | if you don't know what a build server is, you're going to be confused regardless |
| 19:30:15 | dom96 | yeah, but it's meant to provide latest nimrod binaries for new users too |
| 19:30:38 | dom96 | might be nice to get that table into the main nimrod website. |
| 19:30:40 | dom96 | under downloads |
| 19:30:43 | filwit | ahh i see. i didn't realize there was more too it than just a progress report |
| 19:30:54 | filwit | did you make this Dom? |
| 19:30:57 | dom96 | yeah |
| 19:31:03 | filwit | it's nice, now that i take a second look |
| 19:31:12 | filwit | i like your icons and layout |
| 19:31:28 | dom96 | it's a bit all over the place |
| 19:31:28 | filwit | (the documentation/C sources icons) |
| 19:31:40 | filwit | now it's good |
| 19:31:45 | filwit | no** it's good |
| 19:31:56 | dom96 | I took those icons off some website :P |
| 19:32:22 | filwit | one thing I might do though, is make the Documentation/CSources buttons a tad bigger, to make them stand out |
| 19:32:26 | dom96 | and now that I look at the documentation icon all I see is a fuel station sign :P |
| 19:32:48 | filwit | really? i see a book |
| 19:33:06 | filwit | but yeah, lol, i can see the gasstation now |
| 19:33:11 | dom96 | heh |
| 19:33:50 | filwit | okay, g2g |
| 19:33:52 | filwit | later |
| 19:33:55 | dom96 | see you later! |
| 19:33:56 | * | filwit quit (Quit: Leaving) |
| 19:33:56 | Araq | bye |
| 19:43:55 | dom96 | Araq: You wanna have NimBot announce new issues and pull requests from github? |
| 19:45:37 | Araq | dom96: sure |
| 19:48:34 | * | zahary quit (Quit: Leaving.) |
| 20:12:03 | * | Zerathul joined #nimrod |
| 20:22:54 | _ponce | we use lots of SIMD at work for trancoding, and best results are more often than not loop with complier intrinsics rather than inline assembly or pure C++ with some pragmas |
| 20:23:09 | _ponce | loops with compiler intrinsics* |
| 20:23:48 | _ponce | disambiguating pointer aliasing sometimes help |
| 20:24:40 | Araq | making the backend generate 'restrict' is on my todo |
| 20:24:56 | Araq | I'm not sure if it should be the default though |
| 20:25:08 | _ponce | with ICC, restrict rarely lends to gains (!) |
| 20:25:21 | _ponce | mightbe useful for GCC though |
| 20:25:52 | Araq | ICC often generates 2 versions and determines aliasing at runtime |
| 20:26:07 | Araq | at least it used to do that ... |
| 20:26:26 | _ponce | didn't know that |
| 20:27:47 | _ponce | http://www.unix.com/man-page/all/9f/memmove/ < restrict is not a reasonable default |
| 20:28:34 | Araq | afaik it's the default in Fortran |
| 20:28:57 | Araq | and often the code is not written with aliasing in mind anyway |
| 20:29:27 | Araq | I'd guess most programmers never think of it and assume there is no aliasing |
| 20:30:54 | gour | Araq: pong |
| 20:31:15 | Araq | gour: it's not enough to use #def only once for that header :P |
| 20:31:31 | Araq | you need to tell c2nim what FAR means (nothing) etc. |
| 20:32:00 | gour | hmm, so s/define/def/g? |
| 20:32:10 | Araq | worth a try, yeah |
| 20:32:57 | gour | ok. will try...a bit later when we finish some afk work...in any case i've to make this work ;) |
| 20:34:57 | gour | btw, is there any work planned for nimrod to make it easier to do C++ bindings (e.g. Qt & wx)? |
| 20:39:22 | Araq | sure but it's not of high priority |
| 20:39:50 | Araq | on the other hand, parsing C++ is a challenge |
| 20:41:13 | Araq | so I may do it as soon as ambition gets me again ;-) |
| 20:41:32 | gour | i know...C++ bindings are hard for every language...otoh, today i read that even gnucash devs are considering (long-term) to move to C++ bindings not being overly happy with the direction of gtk+ project |
| 20:41:46 | * | gour will pray for Araq'a ambitions :-) |
| 20:46:13 | Araq | gour: http://nimrod-code.org/nimrodc.html#importcpp-pragma |
| 20:48:43 | gour | hmm..wonder if it would be good-enough for projects of qt/wx size |
| 22:16:43 | gour | dom96: have you forgot to comment here http://www.reddit.com/r/programming/comments/18r7zk/a_mathematician_looks_at_d/ ? |
| 22:18:22 | dom96 | nah, I just decided not to. |
| 22:18:42 | dom96 | most people will read my comment on the site anyway |
| 22:19:09 | gour | how to comment on the site? |
| 22:19:56 | dom96 | you need to register on it |
| 22:20:03 | dom96 | why, what do you wanna say? |
| 22:20:41 | gour | that i've chosen nimrod amongst several other langs finding it just the 'right' langu |
| 22:20:45 | gour | *langauge |
| 22:21:12 | dom96 | cool |
| 22:21:14 | * | shevy joined #nimrod |
| 22:36:19 | gour | dom96: posted |
| 22:37:54 | gour | but the comments are moderated :-/ |
| 22:38:35 | dom96 | yeah :\ |
| 22:39:19 | Araq | yeah, spam the poor guy with nimrod advertising |
| 22:40:01 | gour | saving his soul from entanglement with D |
| 22:40:04 | gour | :-) |
| 22:42:12 | Araq | btw I finally replied to your post |
| 22:43:13 | gour | thanks. i saw it...i'll also think about tup...which seems to be good for incremental builds, although not sure how/if it is relevant for nimrod |
| 22:43:38 | Araq | nimrod can do incremental builds on his own |
| 22:43:49 | Araq | well ... ok, currently it's broken |
| 22:44:06 | Araq | but the technology is there |
| 22:44:07 | gour | heh...i'm thinking speed-wisw |
| 22:44:11 | gour | *wise |
| 22:45:10 | gour | otoh, it seems that there is technology to do doc-building (reST --> html/pdf), so not much is missing, for my use-case |
| 22:46:24 | gour | btw, there was one guy visiting this place some time ago, liking nimrod, but complaining about the license. not here any longer? |
| 22:48:00 | Araq | not here anymore, but I know where to find him and he'll still use it once the licence changed |
| 22:49:13 | gour | so much occupied with it? |
| 22:50:39 | Araq | *shrug* it's his religion |
| 22:50:47 | gour | :-) |
| 23:16:33 | * | gour quit (Quit: WeeChat 0.4.0) |
| 23:28:52 | * | Zerathul quit (Quit: ChatZilla 0.9.90 [Firefox 18.0.2/20130201065344]) |